An Empirical Study on Herd Behavior in Cryptocurrency Trading

In recent years cryptocurrency trading has been very active and received attention from worldwide investors. Given its short history and lack of effective regulatory frameworks, cryptocurrency trading is still like the wild west. This study selected the top 21 dominant cryptocurrencies to analyze the trading behavior associated with them over the period from 2016 to May 2023. We adopted the cross-sectional absolute deviations (CSAD) model to investigate investors' trading behavior. Two calculation methods are used: the capitalization-weighted method and the equally weighted method. Our analysis results have indicated that no matter which calculation method is used, the cryptocurrency marketplace has shown strong and significant herd behavior from 2016 to March 12, 2020. However, significant herd behavior couldn't be detected after March 12, 2020. This is a very interesting and yet important finding from our study. The implication is that March 12, 2020 appeared to be a turning point where important market conditions might have changed, such as the dominant cryptocurrency trading force has transformed from individuals to institutional investors.